How can I reduce the time for simulation using this model? [FEK0 2023.1]

sswh2

feko_bending_aps_KD_fixed_a_decrease(perfect_layer).out

Hello, my simulation with general MOM solver is taking more than four hours when I run this design with only 30 number of points, I tried MLFMM to run it faster but MLFMM is showing "feko internal error", see the attached design model of CADFEKO and the .out file for MLFMM error , kindly tell me the way how can I run this model in between one hour to check my result?

feko_bending_aps_KD_fixed_a_decrease(perfect_layer_parametric).cfx

Zeina

From the .out file you have shared, it seems that you are not using parallel runs:
NOTE 35127: You are running a sequential version of Altair Feko, but it was detected that your machine has more than one core. For best performance, it would be recommended to run the parallel Feko Solver

In order to accelerate the runs you can start by allowing more cores to be used (Sove/Run tab activate Parallel)

On another hand, switching to MLFMM in this case does not seem to be a good idea, again, as mentioned in the .out:
WARNING 32945: It is highly recommended to use the conventional MoM, since the MLFMM near field matrix fill exceeds 20% of the full MoM matrix

Torben Voigt

Hi Srabonty,

Again you chose mesh of lambda/50. This is why it takes so long to run. I told you twice already ;)

Please use standard mesh, maybe with refinements along the patch edges.

Best regards,
Torben

Zeina

From the .out file you have shared, it seems that you are not using parallel runs:
NOTE 35127: You are running a sequential version of Altair Feko, but it was detected that your machine has more than one core. For best performance, it would be recommended to run the parallel Feko Solver

In order to accelerate the runs you can start by allowing more cores to be used (Sove/Run tab activate Parallel)

On another hand, switching to MLFMM in this case does not seem to be a good idea, again, as mentioned in the .out:
WARNING 32945: It is highly recommended to use the conventional MoM, since the MLFMM near field matrix fill exceeds 20% of the full MoM matrix

Torben Voigt

Hi Srabonty,

Again you chose mesh of lambda/50. This is why it takes so long to run. I told you twice already ;)

Please use standard mesh, maybe with refinements along the patch edges.

Best regards,
Torben

sswh2

@Torben Voigt sir Thanks, Now I will run with standard mesh with same wire radius and will mention here if it takes same amount of time or not. Let's see. Thanks.

sswh2

@Torben Voigt sir thanks a lot, you were right standard meshing takes 14 min to run it, I thought the result should be changed if I don't apply lambda/50 in meshing, but result is also a good little bit better than the previous mesh case.